Contribution of intraoperative electroencephalogram suppression to frailty-associated postoperative delirium: mediation analysis of a prospective surgical cohort

Abstract

BackgroundFrailty is a risk factor for postoperative delirium (POD), and has led to preoperative interventions that have reduced, but not eliminated, the risk. We hypothesised that EEG suppression, another risk factor for POD, mediates some of the frailty risk for POD. MethodsA prospective cohort study enrolled patients aged 65 yr or older, scheduled for noncardiac surgery under total intravenous anaesthesia. Frailty was assessed using the FRAIL scale. Cumulative duration of EEG suppression, defined as an amplitude between –5 and 5 μV for >0.5 s during anaesthesia, was measured. POD was diagnosed by either confusion assessment method (CAM), CAM-ICU, or medical records. The severity of POD was assessed using the Delirium Rating Scale – Revised-98 (DRS). Mediation analysis was used to estimate the relationships between frailty, EEG suppression, and severity of POD. ResultsAmong 252 enrolled patients, 51 were robust, 129 were prefrail, and 72 were frail. Patients classified as frail had higher duration of EEG suppression than either the robust (19 vs 0.57 s, P<0.001) or prefrail groups (19 vs 3.22 s, P<0.001). Peak delirium score was higher in the frail group than either the robust (17 vs 15, P<0.001) or prefrail groups (17 vs 16, P=0.007). EEG suppression time mediated 24.2% of the frailty–DRS scores association. ConclusionEEG suppression time mediated a statistically significant portion of the frailty–POD association in older noncardiac surgery patients. Trials directed at reducing EEG suppression time could result in intraoperative interventions to reduce POD in frail patients. Clinical trial registrationChiCTR2000041092 (Chinese Clinical Trial Registry).